执行算法递归地寻找较低的数字，非常慢

Question

Hi I have the following code

#include <stdio.h>
#include <stdlib.h>
#include <time.h>
#include <string.h>

#define min(x, y)(x < y)?(x):(y)
#define SIZE 1000

int valormenor(int a[], int n)
{
    if(n == 1)
        return a[0];
    else
        return min(a[0], valormenor(a + 1, n - 1));
}

int main(void)
{
    int arr[SIZE] = {0}, i;
    srand (time (NULL));
    for(i = 0; i < SIZE; i++)
        arr[i] = rand() % SIZE;
    arr[5] = -1;
    printf("%d\n", valormenor(arr, SIZE));

    return 0;
}

The point is that do not understand because it takes too long to find the smallest number, my theory is that this recursive function is badly implemented, you who claim to?

Answer 1

Let's expand the min macro here:

return min(a[0], valormenor(a + 1, n - 1));

That becomes

return (a[0] < valormenor(a + 1, n - 1))?(a[0]):(valormenor(a + 1, n - 1));

As you can see, valormenor is called twice. Those two recursive calls make four recursive calls, which make eight recursive calls, and so on. It's a classic double evaluation bug.

Don't use macros like this. They're just not worth the headaches.